lyrics_grabber.py

#!/usr/bin/env python
 
from pyquery import PyQuery as pq
import re
from unidecode import unidecode
import optparse
 
def get_lyrics(band):
    band = re.sub(r'\s','_',band)
    url = "http://www.allthelyrics.com/lyrics/" + unidecode(band)
    bandpage = pq(url=url)
    bandpage.make_links_absolute()
 
    for song in bandpage(".lyrics-list-item a"):
        song_url = pq(song).attr("href")
        lyrics = pq(url=song_url)
        for lyric in lyrics(".content-text-inner"):
            print re.sub(r'<[^>]+>',"",unidecode(pq(lyric).html()));
 
 
def main():
    p = optparse.OptionParser()
    p.add_option("-b","--band",
            help="Name of the Band to get lyrics from");
    options, arguments = p.parse_args()
    band_name = options.band
    get_lyrics(band_name)
 
if __name__ == '__main__':
    main()